Domain Dependent Optimal Pattern Generation
Title: Domain Dependent Optimal Pattern Generation
SNIC Project: SNIC 2021/22-825
Project Type: SNIC Small Compute
Principal Investigator: Paul Höft <paul.hoft@liu.se>
Affiliation: Linköpings universitet
Duration: 2021-10-21 – 2022-11-01
Classification: 10201
Keywords:

Abstract

We will discover, evaluate and compare optimal pattern collections on standardized domain suites to obtain domain dependent optimal pattern generation algorithms. This allows for the classification of planning problems into complexity classes based on the pattern sizes. Additionally we plan to use this as training data to learn domain independent pattern generation algorithms as a follow up work.